The media, and vendor marketing departments, have described RPA as a compelling technology for enterprises, based on its perceived ability to reduce operational cost. However, our review of case studies and BPS vendor offerings reveals a more complex set of drivers at work. Here I take a brief look at examples of the application of RPA in the banking sector.

The banking industry has been changing over the last eight years along four key dimensions:

• Distribution system: primarily a reduction in physical channels (i.e. branches) and increase in electronic channels (i.e. web, mobile, etc.)
• Customer base:
  • Mature customers: primarily older customers in mature economies who are knowledgeable about financial services. Maintaining these customers requires a shift from product-centric delivery to a customer-centric approach based on lifecycle management
  • New customers: primarily younger customers in mature economies, and unbanked in emerging economies, requiring omni-channel digital capabilities
• Offerings: shift from risk-based products (i.e. loans) to transaction-based products (i.e. payments, wealth management, and deposits)
• Compliance: increased compliance has necessitated improved process accuracy and transparency.

Each of these industry changes requires increased investment in process change, but are subject to:

• Flat or declining revenues from lower risk offerings, with attendant lower revenue/profit margin per product
• Lower revenue per customer, as customers new to banks are (statistically) lower income customers

Under these circumstances, overall cost is not as important as cost per unit of delivery. The old model of high fixed cost, offset by very low marginal cost, does not work when high volumes are not possible or are transient. Banks need greater flexibility to manage operational delivery under conditions of greater transaction cyclicality and product obsolescence.

Vendors delivering RPA within bank operations have identified multiple drivers for RPA adoption. For example, Tech Mahindra identifies three key use cases for RPA in banking:

• Cost:
  • Faster execution: 3 to 4 times faster execution than typically with humans
  • Improved quality: fewer mistakes and greater STP
  • Longer processing windows: ability to work 24 hours per day
  • Improved employee effectiveness: repetitive tasks offloaded from humans, allowing them to focus on complex decision-based tasks
• Growth:
  • Repurposed bots: Bots can be reassigned to differing tasks when workloads decline cyclically
  • Extended windows: As volumes scale, bots can extend working windows as required at no additional cost. If additional capacity is required, new bots can be added or the workload prioritized
  • Market customization: Bots can be programed to execute processes with market variations easily, and for more markets than a human can effectively master
• SLAs and KPIs:
  • Compliance: Bots can be programed to execute compliance processes with minimal transition costs (no training classes) and high accuracy
  • Accuracy: lack of human error in execution (conversely, inaccurately programed bots will repeat errors at high speed)
  • Data monitoring: Bots produce operational data in high quantity, which can be used to deliver additional process improvements.

The result is that the application of RPA technology is allowing banks to enter markets, and deliver offerings which were not attractive in the past due to low anticipated volumes, profit margins, or cyclicality. The ability of banks to pursue previously unattractive business opportunities is allowing them to diversify their businesses across a broader set of opportunities. RPA makes this possible by creating more flexible operations delivery capabilities which can adapt to heterogeneous requirements, without confusion or fatigue from drudgery.